EIGRP Equal Cost Load Balancing IPV4 IPV6

EIGRP Equal Cost Load Balancing

This article is on EIGRP interface bandwidth configuration and how to configure EIGRP Equal Cost Load Balancing. This article will also give to a short introduction of EIGRP hello and hold down timer.

EIGRP Bandwidth Configuration For IPv4

By default, EIGRP uses only up to 50% of the bandwidth of an interface for EIGRP information. This prevents the EIGRP process from overusing the links and not allowing enough bandwidth for normal traffic routing.
Use the ip bandwidth-percent eigrp command to configure the percentage of bandwidth that EIGRP can use in an interface.
Router (config-if) # ip bandwidth-percent eigrp  as-number 
In Image, R1 and R2 share a very slow 64 kb / s link.

The configuration to limit the bandwidth used by EIGRP is shown below.
R1 (config) # interface serial 0/0/0
R1 (config-if) # ip bandwidth-percent eigrp 1 40
R1 (config-if) #
R2 (config) # interface serial 0/0/0
R2 (config-if) # ip bandwidth-percent eigrp 1 40
R2 (config-if) # The ip bandwidth-percent eigrp command uses the configured bandwidth (or the default bandwidth) to calculate the percentage that EIGRP can use. In this example, EIGRP is limited to no more than 40% of the link bandwidth. Therefore, EIGRP never uses more than 32 kb / s of link bandwidth for EIGRP packet traffic. To restore the default value, use version no of this command.

IPv6 EIGRP Bandwidth Configuration

To configure the percentage of bandwidth that EIGRP can use for IPv6 on an interface, use the ipv6 bandwidth-percent eigrp command in interface configuration mode. To restore the default value, use version no of this command. The following shows the configuration of the interfaces between R1 and R2 to limit the bandwidth used by EIGRP for IPv6.

R1 (config) # interface serial 0/0/0
R1 (config-if) # ipv6 bandwidth-percent eigrp 2 40
R1 (config-if) #
R2 (config) # interface serial 0/0/0
R2 (config-if) # ipv6 bandwidth-percent eigrp 2 40
R2 (config-if) #

Hello Timer & holddown timer for EIGRP IPV4

EIGRP uses a lightweight greeting protocol to establish and control the connection status of neighbors. The timeout tells the router the maximum amount of time to wait to receive the next greeting, before declaring that the neighbor is unreachable. Hello Timer and timeouts can be configured per interface and do not have to match other EIGRP routers to establish or maintain adjacencies. The command to configure a different greeting interval is as follows:
Router (config-if) # ip hello-interval eigrp  as-number seconds
If you change the greeting interval, make sure that the timeout value is equal to or greater than the greeting interval. Otherwise, neighbor adjacency is deactivated after the timeout expires and before the next greeting interval. To set a different timeout, use the following command:
Router (config-if) # ip hold-time eigrp  as-number secondsThe value of seconds for the greeting and timeout intervals can be from 1 to 65,535. The R1 configuration is shown below to use a 50 second greeting interval and a 150 second timeout. You can use the form no of the two commands to restore the default values.
R1 (config) # interface s0 / 0/0
R1 (config-if) # ip hello-interval eigrp 1 50
R1 (config-if) # ip hold-time eigrp 1 150It is not necessary for the greeting interval time and the waiting time to match for two routers to form an EIGRP adjacency.

Hello Timer  and holddown timer for EIGRP IPV6

EIGRP for IPv6 uses the same hello and wait interval times as EIGRP for IPv4. The interface configuration mode commands are similar to those used for IPv4:
Router (config-if) # ipv6 hello-interval eigrp  as-number seconds
Router (config-if) # ipv6 hold-time eigrp  as-number secondsNext, the configuration of the greeting and waiting interval times for R1 and R2 with EIGRP for IPv6 is shown.
R1 (config) # inter serial 0/0/0
R1 (config-if) # ipv6 hello-interval eigrp 2 50
R1 (config-if) # ipv6 hold-time eigrp 2 150
R2 (config) # inter serial 0/0/0
R2 (config-if) # ipv6 hello-interval eigrp 2 50
R2 (config-if) # ipv6 hold-time eigrp 2 150

IPV4 EIGRP Equal Cost Load Balancing 

Load balancing of equal cost is the ability of a router to distribute outbound traffic across all interfaces that have the same metric from the destination address. Load balancing uses network segments and bandwidth more efficiently. In the case of IP, Cisco IOS software applies load balancing by default with up to four routes of equal cost.
In Image shown below is a EIGRP network topology.

In this topology, R3 has two EIGRP routes of equal cost to the network between R1 and R2, 172.16.3.0/30. One route is through R1 at 192.168.10.4/30, and the other is through R2 at 192.168.10.8/30.

LOAD BALANCING EXAMPLE

The show ip protocols command can be used to verify the amount of routes of equal cost that are currently configured on the router. The result of Image 4 shows that R3 uses the default option of four routes of equal cost.

The routing table maintains both routes. In Image 5, it is shown that R3 has two EIGRP routes of equal cost for the 172.16.3.0/30 network. One route is through R1 in 192.168.10.5, and the other is through R2 in 192.168.10.9.

If the topology of Image 3 is observed, it may seem that the route through R1 is the best route, because there is a 1544 kb / s link between R3 and R1, while the link to R2 is only 1024 kb / s
However, EIGRP only uses the slowest bandwidth in the composite metric, which is the 64 kb / s link between R1 and R2. The two routes have the same 64 kb / s link as the slowest bandwidth, which results in both routes being equal. When process switching is applied to a package, load balancing of routes of equal cost is produced per package. When fast switching is applied to the packets, the load balancing of routes of equal cost is produced by destination. Cisco Express Forwarding (CEF) can perform load balancing by package and destination.

THE MAXIMUM-PATHS COMMAND

By default, the Cisco IOS allows load balancing to use up to four routes of equal cost. However, this can be modified. With the maximum-paths command of the router configuration mode, up to 32 routes of equal cost can be maintained in the routing table.
Router (config-router) # maximum-paths  value
The value argument refers to the number of routes that must be maintained for load balancing. If the value is set to 1 , load balancing is disabled.

IPV6 EIGRP Equal Cost Load Balancing 

In Image 6, the EIGRP network topology for IPv6 is shown. Serial links in the topology have the same bandwidth that is used in the EIGRP topology for IPv4.

Similar to the previous situation for IPv4, R3 has two EIGRP routes of equal cost for the network between R1 and R2, 2001: DB8: CAFE: A001 :: / 64. One route is through R1 on FE80 :: 1, and the other is through R2 on FE80 :: 2.
In following Figure, it is shown that the EIGRP metrics for the 2001: DB8: CAFE: A001 :: / 64 and 172.16.3.0/30 networks are the same in the IPv6 routing table and in the IPv4 routing table.

This is because the composite metric of EIGRP is the same in EIGRP for IPv6 and for IPv4.

LOAD BALANCING WITH DIFFERENT COST

EIGRP for IPv4 and for IPv6 can also balance traffic on several routes with different metrics. This type of balancing is called “ load balancing with different costs ”. Setting a value with the variance command in router configuration mode allows EIGRP to install multiple routes without loops and at a different cost in a local routing table.

• To be installed in the local routing table, a route that is discovered through EIGRP must meet two criteria:
• The route must not have loops and must be a feasible successor or have a reported distance less than the total distance.
• The route metric must be less than the best route metric (the successor) multiplied by the variation configured on the router.

For example, if the variation is 1, only routes with the same metric as the successor are installed in the local routing table. If the variation is 2, any route discovered by EIGRP with a metric less than double the successor's metric is installed in the local routing table.
To control how traffic is distributed between routes when there are several routes with different costs for the same destination network, use the traf fi c-share balanced command . Traffic is distributed proportionally according to the proportion of costs.